extropian
Re:Show HN: Claude Code Usage Monitor – real-time tracker to dodge usa...
on Friday June 20, 2025 @01:42AM (#44325042)
Keep an eye on your memory usage after installing this extension.
I did a bisect and found this one to be consuming ~1.5gb memory alone and that's when I removed it. YMMV.
Re:From ChatGPT to Temu: ranking top Internet services in 2024
on Thursday January 02, 2025 @10:20AM (#42575170)
Facebook solidly at #2 is surprising, considering they've been heavily optimising the feed for algorithmic short video content like Instagram and Tiktok.
Are people using it for specific services like groups & marketplaces - or do people still use it to keep up with their friends?
Re:Most people don't care about quality
on Thursday January 02, 2025 @10:10AM (#42575059)
Most people do care about quality. There are so few things of great quality that most have to remain content with the available options.
Like in the article, anyone watching Netflix's badly written shows have probably watched all the great ones there - and there are not a lot of great ones to go through in most niches.
Re:AI and "Web 3.0" (2006)
on Monday November 25, 2024 @11:26PM (#42242597)
"These early applications work from detecting patterns in data. They have no understanding of language. They cannot reason about user goals. They have no base of knowledge that would allow them to make common sense connections."
^A prophetic quote, one that's still used to dismiss LLMs today.
Re:Ask HN: How are you using LLMs?
on Monday November 25, 2024 @10:35PM (#42242354)
Recently, I've been enjoying just talking and brainstorming with Claude. It feels almost like talking to a really smart friend, and that hasn't been replicated by any other model.
It's a shame, because GPT-4o, despite being higher on the leaderboards should've been able to do the same, but it's not even close how robotic and LLM-esque(?) the conversations feel, compared to Claude.